我正在尝试使用Camel路由器以5分钟的周期间隔将消息从queue1(死信队列)移动到活动MQ中的queue2。我使用下面的代码来实现这个目标: -
public class MessageRouteBuilder extends RouteBuilder {
private static final Logger LOG =
LoggerFactory.getLogger(MessageRouteBuilder.class);
/*
* (non-Javadoc)
*
* @see org.apache.camel.builder.RouteBuilder#configure()
*/
@Override
public void configure() throws Exception {
LOG.info("Routing of camel is started");
CronScheduledRoutePolicy startPolicy = new CronScheduledRoutePolicy();
startPolicy.setRouteStartTime("0 0/5 * * * ?");
from(
"jms:queue:DLQ.Consumer.OUTDOCS.VirtualTopic.queue1")
.routeId("DLQMessageMoverID").routePolicy(startPolicy)
.noAutoStartup()
.to("jms:queue:Consumer.OUTDOCS.VirtualTopic.queue1");
LOG.info("Routing of camel is done");
}
}
@Startup
@Singleton
public class ScheduledMessageDLQConsumer {
@Inject
private MessagingUtil msgUtil;
@Inject
private MessageRouteBuilder builder;
private static final Logger LOG =
LoggerFactory.getLogger(ScheduledMessageDLQConsumer.class);
@PostConstruct
public void init() {
LOG.info("camel Scheduling scheduled started");
CamelContext camelContext = new DefaultCamelContext();
ConnectionFactory connectionFactory = msgUtil.getAMQConnectionFactory();
camelContext.addComponent("jms", JmsComponent.jmsComponentAutoAcknowledge(connectionFactory));
try {
camelContext.addRoutes(builder);
camelContext.start();
LOG.info("Camel scheduling completed");
} catch (Exception e) {
// TODO Auto-generated catch block
LOG.error("Error in registering camel route builder", e);
}
LOG.info(" camel Scheduling scheduled completed");
}
}
问题在于: - 5分钟后启用Camel路由。它将消息从DLQ(DLQ.Consumer.OUTDOCS.VirtualTopic.queue1)移动到queue1(Consumer.OUTDOCS.VirtualTopic.queue1)。但是如果消息是有毒的,它又会回到DLQ并再次路由将消息从DLQ移动到正常队列,这个过程继续无限运行。
我的要求是路由应该每隔5分钟将消息从DLQ移到队列一次?如果有毒信息出现,应该只在5分钟后检查。
答案 0 :(得分:1)
首先,你的整个想法看起来都是糟糕的设计。重新处理和重新传递应该在消费者或经纪人处理,没有任何模糊的期刊" DLQMessageMover"。如果您正在控制 OUTDOCS.VirtualTopic.queue1 使用应用程序,请重新考虑错误处理的概念。
BTW在消费者连接上的maximumRedeliveries = - 1和redeliveryDelay = 300000的简单组合与所有代码具有相同的影响,您已在此问题中写过。
其次,您需要idempotent consumer在标头上使用名称为JMSCorrelationID的相关键。此过程每个相关ID只有一次。使用MemoryIdempotentRepository时,它会在路由重启时被清除,因此会再次处理消息,这符合您的要求。
我创建了一个小例子来展示它是如何工作的。在您的情况下,将不会模拟JMSCorrelationID标头和jms组件而不是计时器。
public class IdempotentConsumerRouteBuilder extends RouteBuilder {
private final IdempotentRepository idempotentRepository = new MemoryIdempotentRepository();
private final List<String> mockCorrelationIds = Arrays.asList("id0","id0","id0","id1","id2","id0","id4","id0","id6","id7");
public void configure() {
CronScheduledRoutePolicy startPolicy = new CronScheduledRoutePolicy();
startPolicy.setRouteStopTime("0 0/5 * * * ?");
startPolicy.setRouteStartTime("0 0/5 * * * ?");
from("timer:jms?period=100")
.routePolicy(startPolicy)
.process(e -> e.getIn().setHeader(
"JMSCorrelationID", //Mock JMSCorrelationID to work with timer as it is jms component
mockCorrelationIds.get(e.getProperty("CamelTimerCounter", Integer.class)%10))
)
.idempotentConsumer(header("JMSCorrelationID"), idempotentRepository)
.log("correlationId is ${header.JMSCorrelationID}")
.to(("log:done?level=OFF"))
.end();
}}
此代码的输出:
[artzScheduler-camel-1_Worker-3] DefaultCamelContext INFO Route: route1 started and consuming from: timer://jms?period=100
[mel-1) thread #4 - timer://jms] route1 INFO correlationId is id0
[mel-1) thread #4 - timer://jms] route1 INFO correlationId is id1
[mel-1) thread #4 - timer://jms] route1 INFO correlationId is id2
[mel-1) thread #4 - timer://jms] route1 INFO correlationId is id4
[mel-1) thread #4 - timer://jms] route1 INFO correlationId is id6
[mel-1) thread #4 - timer://jms] route1 INFO correlationId is id7
[artzScheduler-camel-1_Worker-6] DefaultShutdownStrategy INFO Starting to graceful shutdown 1 routes (timeout 10000 milliseconds)
[el-1) thread #5 - ShutdownTask] DefaultShutdownStrategy INFO Route: route1 shutdown complete, was consuming from: timer://jms?period=100
[artzScheduler-camel-1_Worker-6] DefaultShutdownStrategy INFO Graceful shutdown of 1 routes completed in 0 seconds
[artzScheduler-camel-1_Worker-6] DefaultCamelContext INFO Route: route1 is stopped, was consuming from: timer://jms?period=100
[artzScheduler-camel-1_Worker-8] ScheduledRoutePolicy WARN Route is not in a started/suspended state and cannot be stopped. The current route state is Stopped
[artzScheduler-camel-1_Worker-7] DefaultCamelContext INFO Route: route1 started and consuming from: timer://jms?period=100
[mel-1) thread #6 - timer://jms] route1 INFO correlationId is id0
[mel-1) thread #6 - timer://jms] route1 INFO correlationId is id1
[mel-1) thread #6 - timer://jms] route1 INFO correlationId is id2
[mel-1) thread #6 - timer://jms] route1 INFO correlationId is id4
[mel-1) thread #6 - timer://jms] route1 INFO correlationId is id6
[mel-1) thread #6 - timer://jms] route1 INFO correlationId is id7
[rtzScheduler-camel-1_Worker-10] DefaultShutdownStrategy INFO Starting to graceful shutdown 1 routes (timeout 10000 milliseconds)